Breast density is known as a strong risk factor for breast cancer. Clinical assessment of breast density during screening mammography is often done by radiologists through visual evaluation or by a computer program. Automated computer methods offer the potential for non-subjective density assessments. With the rapid development and increased utilization of tomosynthesis clinically, there is a practical need for systems to provide automated breast density measurements in tomosynthesis like those available in mammography. QuantraTM is a software package using physical modeling of mammography systems, and performs volumetric assessment of breast tissue composition for conventional mammography. In this paper, we describe recent developments to extend Quantra to calculate breast density using tomosynthesis projection images. Our development took advantage of the combo imaging mode of Hologic Selenia DimensionsTM system, which allowed co-registered conventional 2D mammogram and 3D tomosynthesis images to be acquired in a single compression. We used the Quantra results of 2D mammograms as a reference to refine the new processing algorithm for tomosynthesis images. This paper describes details of the new algorithm and provides some preliminary results.
[1]
J A Seibert,et al.
An edge spread technique for measurement of the scatter-to-primary ratio in mammography.
,
2000,
Medical physics.
[2]
Ann-Katherine Carton,et al.
Breast percent density estimation from 3D reconstructed digital breast tomosynthesis images
,
2008,
SPIE Medical Imaging.
[3]
J M Boone,et al.
Scatter/primary in mammography: Monte Carlo validation.
,
2000,
Medical physics.
[4]
J. Boone,et al.
Scatter/primary in mammography: comprehensive results.
,
2000,
Medical physics.
[5]
J A Seibert,et al.
X-ray scatter removal by deconvolution.
,
1988,
Medical physics.